The use of HPLC to identify precisely individual retinoids is a methodology that is subject to infinite variability (see, for example, ref. 40). The variables include normal-phase or reverse-phase, type of column, composition of buffers, mix of buffers, single-step or multistep elution, step or gradient elution, and rates of flow. The idea is to use a system such that each of the retinoids one is interested in will come off at a different and highly repeatable elution time. Reverse-phase is generally preferred over normal-phase chromatography, the latter suffering from difficulty in reproducing precise elution times. Therefore, which method should one use?
The method that was used to identify tRA in the chick limb bud by Thaller and Eichele (3) is described below with only minor modifications, since it has been successfully used by several groups, including ourselves, to identify retinoids in the mouse embryo (5). An example of this type of chromatography is given in Fig. 2. Figure 2A shows the separation of a mixture of standards and Fig. 2B, the retinoids present in a mouse embryo. However, this method was developed before the discovery of 9-cis-RA, and it fails to separate this isomer because 9-cis coelutes with retinol, so this is by no means a perfect method. To identify 9-cis, we have developed a method (18) that involves first running the sample on a normal-phase system with an NH2 column that retains only the acids and then running the acid fractions on a reverse-phase system as described below.
Because there is no contaminating retinol left, all the tRA isomers can be detected, but one does not obtain any quantitation for retinol by this method. However, Kraft et al. (12) describe a method involving a gradient elution of a mixture of ammonium acetate and methanol whereby 9-cis is identifiable as a separate peak.
Most embryological interest has been centered on the retinoic acids and retinol, but there are other retinoids of biological interest, such as the retinyl esters, that these methods do not identify, because they come off the column after very long elution times. Another method that was designed to separate 14 different retinoids, including those of greatest interest, such as 4-oxo-RA, tRA, and retinol, as well as retinyl acetate, palmitate, and stearate, was described by Cullum and Zile (41). This involves a multistep, three solvent (water, methanol, chloroform) gradient system eluting for 50 min. Thus, the choice of method is dependent on which retinoids one wants to identify.
Was this article helpful?